Loss to Follow-Up, Improvement in Oral Health–Related QoL, and Self-Reported Oral Health Complaints by Intervention, Excluding GP Offices With Follow-Up <60%
Experimental | Control | χ2 Statistic | P Value | Intracluster Correlation Coefficient | Adjusted χ2 Statistic | Adjusted P Value | |
---|---|---|---|---|---|---|---|
Study details | |||||||
No. of GP offices | 8 | 10 | |||||
No. of study participants | 210 | 318 | |||||
Loss to follow-up | |||||||
GP offices | 0 (0) | 0 (0) | N/A | N/A | N/A | N/A | N/A |
Study participants | 39 (18.6) | 48 (15.1) | 1.111 | .292 | 0.067 | 0.356 | .551 |
Improvement in oral health–related QoLa | n = 120b | n = 205b | |||||
Functional limitation | 17 (14.2) | 17 (8.3) | 2.788 | .095 | –0.005c | 2.788 | .095 |
Physical pain | 37 (30.8) | 43 (21.0) | 3.964 | .046 | –0.005c | 3.964 | .046 |
Psychologic discomfort | 21 (17.5) | 20 (9.8) | 4.118 | .042 | –0.008c | 4.118 | .042 |
Physical disability | 12 (10.0) | 12 (5.9) | 1.903 | .168 | 0.033 | 1.152 | .283 |
Psychologic disability | 22 (18.3) | 17 (8.3) | 7.226 | .007 | 0.007 | 6.355 | .012 |
Social disability | 4 (3.3) | 8 (3.9) | 0.069 | .793 | –0.018c | 0.069 | .793 |
Handicap | 15 (12.5) | 10 (4.9) | 6.193 | .013 | 0.010 | 5.173 | .023 |
OHIP-NL14 total score | 46 (38.3) | 51 (24.9) | 6.545 | .011 | –0.005c | 6.545 | .011 |
Improvement in self-reported oral health complaints | n = 171b | n = 270b | |||||
Pain in mouth | 26 (15.2) | 28 (10.4) | 2.277 | .131 | 0.011 | 1.764 | .184 |
Dry mouth | 26 (15.2) | 56 (20.7) | 2.120 | .145 | –0.007c | 2.120 | .145 |
Bad breath | 16 (9.4) | 16 (5.9) | 1.831 | .176 | –0.030c | 1.831 | .176 |
Any | 59 (34.5) | 85 (31.5) | 0.436 | .509 | –0.009c | 0.436 | .509 |
GP = general practitioner; N/A = not applicable because GP office is cluster level itself; OHIP-NL14 = 14-item Oral Health Impact Profile; QoL = quality of life.
Note: Data in Experimental and Control columns from Loss to follow-up down are presented as n (%).
↵a Subdomains according to Slade.21
↵b Number of patients with a completed baseline and follow-up questionnaire.
↵c To calculate adjusted χ2 statistics, negative intracluster correlation coefficients were set to 0 for these variables.30,31